Subscribe
  • Home
  • /
  • CX
  • /
  • Serious about Web chat

Serious about Web chat

Either commit to a full-blown Web chat deployment, or don't deploy the solution at all.

Johan Jacobs
By Johan Jacobs, Independent analyst & executive consultant
Johannesburg, 19 Jan 2015

Web chat is a fast-growing communication channel with mature technology. Unfortunately, most Web chat interactions are sliding down the poor customer experience hill fast; it is becoming a useless contact centre channel - not unlike the phone.

The business case for Web chat is mostly motivated by (1) a reduction in the cost of phone-based interactions; or (2) introducing new interaction channels to the customers. A typical Web chat deployment will start as a pilot project, and becomes the first mistake companies make.

Most pilot projects are not seen as strategic, and as such, a low-cost Web chat solution is selected, which inevitably becomes a permanent installation, often leaving a platform that cannot deliver on all the features that a great customer experience demands. The success of the pilot project will be measured in customer satisfaction and in the TCO of the project. When neither of these delivers on the expectations, Web chat is deemed a failure.

Committed to chatting

The first best practice tip suggests that Web chat pilot projects do not work, so either commit to a full-blown deployment or don't deploy Web chat at all. (Tweet @Bushman10 if this has already happened.)

There are typically two types of costs associated with Web chat: (1) the cost of the technology; and (2) the cost of the agent. Many companies focus only on the former and forget about the opex of the interaction. Typically, the technology costs are in the region of 10% to 20% of that of a contact centre solution, as there is no ACD, PBX, CTI, etc, and cloud solutions bring even more savings, making the business case easy to motivate.

The challenge, however, does not lie in the technology costs, but in the human costs that make up 72% of the cost of a Web chat interaction (28% of interaction cost is technology). So, for real rand-based savings, one needs to focus on the opex of customer interactions and not on the technology costs of the interactions.

The problem with a Web chat interaction is that the duration (handling time) of a Web chat session, on average, is twice as long as that of a phone call. A phone-based agent can service two customers in the same time it takes a Web chat agent to service one customer. The Web chat agent is only half as productive as a phone-based agent, and therefore has twice the TCO as a phone agent, leading to a huge #BusinessCaseFail.

Juggling act

Web chat best practices

1. Either commit to a full Web chat deployment or don't do it at all. Pilot Web chat projects do not work.
2. From the day the company switches on Web chat, agents must perform two to six simultaneous Web chats at a time.
3. Never give phone-based contact centre agents the responsibility to do a Web chat, because they can't type.
4. Do not mix phone calls and Web chat sessions to be dealt with by the same agent at the same time. It is impossible to run multiple chat sessions.

The second best practice tip suggests the very first day Web chat is switched on, the agent must, at a minimum, handle two simultaneous and concurrent Web chat sessions. By engaging with two customers at the same time, the Web chat agent becomes equally productive as a phone-based agent. World Wide Web chat benchmarks, on the other hand, state that in complex Web chat interactions in the financial services and healthcare environments, for example, the average simultaneous Web chat sessions for an agent is four. When performing simplistic customer interactions, the average simultaneous Web chats for an agent is six sessions.

By increasing the number of simultaneous Web chat interactions, a Web chat agent can become twice or three times as effective as a phone-based agent. The 72% of opex costs attributed to a customer interaction can therefore become 36% in complex interactions, or even 24% in easier interactions, and the business case goes from good to great.

There are unfortunately very few agents who can manage so many simultaneous Web chat sessions while delivering an efficient customer interaction.

This brings me to the third best practice tip, which is to never give a phone-based contact centre agent the responsibility to Web chat. (Tweet @Bushman10 if you have made this mistake.)The reason why: call centre agents generally can't type and only use two fingers. The minimum requirement for a Web chat agent is 65 words per minute double-handed typing. If agents can't pass this basic test, then they should not be given the responsibility of a Web chat.

The day a company starts planning the deployment of Web chat, it should also start a new recruitment and training drive to find agents that have both business writing skills and typing skills.

When agents engage in a Web chat session with a customer, they cannot type poorly-constructed responses, nor may they use abbreviations or code, like teenagers texting one another. The responses must be fast and efficient, and the text and content well structured and professional, in order to properly represent the company.

Never give a phone-based contact centre agent the responsibility to Web chat.

It is therefore strongly recommended that in the contact centre, a company creates separate skills groups, where all agents that perform written interaction with customers are separated from agents that perform spoken interactions.

The fourth best practice tip is therefore to never mix phone calls and Web chats with the same agent at the same time. It is impossible to run multiple simultaneous chat sessions and still stream a phone call in between.

In my next Industry Insight, I will explore the fact that Web chat actually has very little to do with Web chat technology and everything to do with knowledge management technology. Keep an eye open for the Industry Insight or tweet me @Bushman10 if you just can't wait.

Share